Composting Corn for Heat Recovery

PROJECT BACKGROUND
My wife and I moved here from Washington State in 1999 and began farming at that time. The home farm has a long history in my family. Currently it is farmed in partnership with my brother. We also rent the farm next door for a total of 260 acres. Corn and soybeans are grown in a 2-year rotation except where mixed grass/alfalfa hay has been included for 2-3 years in the rotation. A small beef cow herd was kept until two years ago. This has permitted rotation of hay around most of the farm, since no hay had previously been grown here for many years. For the past six years strip-tillage has been used for corn and no-tillage for soybeans. Running the rows on the contour along with the waterways we have put in has very significantly reduced the soil erodability and increased the sustainability of our farming operation.

PROJECT DESCRIPTION AND RESULTS
Corn has the potential to be a much less expensive space heating fuel than conventional heating fuels in the Corn Belt. Ethanol is currently produced from the starch fraction of corn grain. The lignocellulosic fraction of corn, which consists of the stover (stalks, cobs, and husks), has limited market potential but contains over half of the energy in the corn plant and is responsible for maintaining high soil organic matter levels upon its return to the soil. Since combustion on a large scale could exacerbate air quality problems and preclude the return of organic matter and nitrogen to the soil, composting may be a preferable non-combustion option for generating heat from corn. This would conserve the nutrients in an organic residue in close proximity to the corn production site for recycling to the soil, maintaining high soil quality. The purpose of this pilot research project conducted with the SARE grant was to recover heat from composting corn grain and stover in a fairly low-tech, small-scale manner for space heating. The project was conducted with consultation from Kapil Arora (Agricultural Engineer, Iowa State University), who has considerable experience in composting.

Problem and Opportunity:
A very substantial fraction of the natural gas, LP, and fuel oil used in the Corn Belt is used simply for space heating. These fuels need to be piped or transported from distant sources to their markets here. Prices for energy and corn have fluctuated wildly recently. But at $1.00 per gallon for LP and $4.00 for corn, there is roughly an equivalent amount of energy in $1.00 worth of each. At $50 per ton for stover, there is about three times as much energy in a dollar’s worth of corn stover than in a dollar’s worth of corn grain or LP at the above prices. Stover has additional allure because it currently has little market appeal otherwise. Using corn as a widely dispersed source of energy for a widely dispersed space heating market has intuitive appeal. Since the ethanol and chemical industries are using primarily carbon, hydrogen, and oxygen, perhaps the fossil fuels currently used for space heating would be better used as transportation fuels and chemicals. Although space heating with corn makes economic sense, its combustion would destroy the organic matter and may preclude any nutrient recovery from the corn. Rapidly composting organic material such as corn stover/grain can maintain temperatures up to 160 degrees F, which are adequate for space heating. If space heating can be done without combustion and the nutrients returned to the soil in organic residue then it would make both economic and ecological sense.

Corn stover is likely to become more marketable in the future for fuel and chemical production. But prices offered to producers for it may leave little profit as compensation for the hazards to soil quality entailed by its removal. Returning large quantities of stover to the soil has been important for soil erosion control, maintaining organic matter levels and good soil quality. Removing part of the stover mass would make it more feasible to eliminate tillage, saving energy and soil organic matter that is destroyed by tillage, while leaving enough cover to protect the soil from erosion. It should also lower the nitrogen fertilizer requirements in continuous corn production. The proposed research would result in a small-scale, low technology way to use corn stover and/or grain by composting it for heat production in a reaction vessel in which optimal conditions for composting are maintained.

The general conditions required for optimal composting and the capacity of composts to generate heat for drying of sludges and other wastes are well established. Much of the composting literature is primarily focused on mass reduction and heat stabilization of waste materials such as sewage sludge, which have a low solids content. Recovery of the heat has not been a focus perhaps because so much of the heat generated is required to evaporate water from wet sludge. Because corn stover and grain are stored at about 15% moisture content, metabolically generated water and water relinquished from mass reduction during the compost process could help maintain compost in an optimal moisture content range of 50 to 60%. Without the need to use the heat generated during composting to evaporate water from the system, some heat might be available for space heating. This is quite a different thermodynamic situation from sludge composting where more than half of the heat generated by composting can be required for drying. Heat has been recovered from composting manure with only modest success. But manure is the fraction of the plant biomass that remains after the animal has extracted the readily available energy. What we proposed in this project was to compost material that was more energy dense than manure, corn stover and grain, for heat production.

The fiber fraction of the corn kernel that remains after ethanol production (distiller’s grains or DDGs) contains all of the plant nutrients that were in the kernel originally. Ethanol is made only from the starch. Livestock producers seek after the high protein coproduct as a protein supplement. But the rapid expansion of the livestock industry that can use large quantities of DDGs has kept pace with the rapid expansion of ethanol production. Although composting DDGs or corn grain for space heating or process heat might provide additional market or utilization opportunities, there have been steep price escalations for both from prices that prevailed at the time this project was proposed. Additionally, the tumbling reactor that was built for this project from a commercial size clothes dryer and equipped to rotate very slowly revealed during preliminary testing that corrosion was going to be a serious problem. This has resulted in more emphasis in this project on static pile composting using stover as the more realistic space heating material.

The initial phase of the project was used to evaluate critical composting parameters in an attempt to maximize or optimize the rate of degradation and heat generation by varying the:
- carbon to nitrogen (C:N) ratio of the mixture
- ratios of grain, stover
- particle size of the components
- aeration rate
- water content
- temperature

Initial Screening Trials:
Rather than congest this report with composting basics and the virtues of compost readers can refer to http://compost.css.cornell.edu/science.html or other sites easily obtained online by doing a search for compost engineering. Likewise, readers interested in references supporting statements made in this report are welcome to request them. The assumptions that are important and specific to the project will be stated as needed. The design of the vessels used for the screening trials and the means of aeration were worked out prior to or very early in the trials. These will be described before discussing the results of the trials along with assumptions about heat loss determination needed to draw conclusions about heat generation. The screening trials were also intended to provide enough information about the variables for composting to preclude the need to spend the entire project budget on instrumentation for measurement of them unless considered to be necessary.

Drum Design:
Several prototypes of the drum reactor were made using 30-gallon plastic drums. The chosen version was made as follows. A large enough opening was cut in the top of the drum to easily add or remove compost material. The top that was removed was perforated with numerous ¼” holes and placed in the bottom of the drum for use as an aeration floor. A 2’ section of 3” diameter galvanized heat duct, sealed at both ends with urethane foam, was placed vertically in the center of the drum with ½” PVC inlet and outlet pipes protruding out the top for air introduction. This was done for heat removal during the trials in case the temperature became too high. A ¾” PVC pipe for compost aeration was placed through the center of the galvanized duct, protruding through the aeration floor at the bottom of the drum and out through the top of the drum lid. A lid consisting of 4” of rigid foam insulation (pinkboard) was fitted over the pipes used for aeration and heat removal and sealed after the compost material was added. A ¾” hole was drilled through the lid so that air could escape from the top of the compost mass during aeration. A smaller hole was drilled for insertion of the compost thermometer. Both holes were stoppered when not used for temperature measurement or aeration.
Compost materials were blended with water in a plastic tub until all of the added water had soaked into the materials before addition to the drums. After adding the compost mixtures and sealing the lids the drums were lined up on a 4” thick floor of rigid foam insulation. The drum walls were wrapped with R-19 fiberglass insulation, which was then wrapped with a sheet of plastic for retention of the insulation against the drum. The insulation value considering the insulation, plastic sheeting, and the plastic drum itself was estimated to be about R-24. The smaller pipes used for heat exchange (never needed) were taped to prevent convective air exchange. The larger pipe used for aeration was connected with vinyl tubing to a manifold for air introduction from an air compressor. The manifold was fitted with a ball valve for each drum. A 30-gallon drum contains a volume of about 4 cubic feet. The 30-gallon drum with the false floor and heat exchanger in place, and allowing for some headspace, contains about 3.25 cubic feet of usable volume.

Aeration:
The targeted aeration rate was based on the rough estimate of 25% degradation of the stover over six weeks with the oxygen requirement based on combustion of the same mass of carbohydrate. A pound of carbohydrate and a pound of oxygen will combust to about 1.4 pounds of carbon dioxide and 0.56 pounds of water. If air is roughly 13 cubic feet per pound and is 21% oxygen the aeration rate should be about 65 cubic feet per pound of degradable material.

A considerable amount of time was spent trying to devise a continuous aeration system that permitted aeration to be measured and controlled independently for each drum. This was important not only for measurement of oxygen supply but also to account for heat loss. Although several strategies were attempted to continually aerate the compost drums, because the volume of air required was so low and the effect of temperature on the pressure regulator was so pronounced a continuous supply of air was too difficult to sustain at a sufficiently constant and measurable rates using the available equipment. Higher airflow rates such as would be used in practice with larger masses of compost should not have presented the same problems. Likewise a CPAP (Constant Pressure Air Pump) as was available later in the study would have worked well. In the end, filling the compressor to its maximum pressure and letting it bleed out entirely delivered a more consistent volume of air. The required volume of air was supplied by repeatedly filling and draining the compressor and regulated for the different aeration rates by opening or closing valves to each drum. This was inferior to a continuous supply of air that would allow more time for oxygen diffusion into micropores of the compost mass, but was a practical compromise.

To obtain some feedback about the oxygen status of the air exhausted during aeration a low flame from a propane torch was placed over the exhaust port to see whether there was sufficient oxygen left in the air to keep the flame lit. Although this worked fairly well in a rough sense it told nothing about the degree of oxygen penetration into the micropores of the compost that occurred during aeration. This technique was not possible to use in the larger piles because the exhaust occurred over a diffuse area.
The screening trials conducted in the drums consisted of four trials. All treatments are unreplicated, so apparent differences should be taken with a grain of salt. Since this is a pilot study done mainly for indications of potential efficacy, replication was sacrificed so that more treatments could be included. Underestimation of the complexity of the project (a.k.a. biting off more than you can chew) also contributed to the early subjugation of the tumbling reactor.

Trial 1 was done partly as a means of testing aeration methods and partly as a means of making an initial determination of the magnitude of response to additions of corn grain and urea to corn stover. Drums were aerated at a constant rate once each day with about 5 cu ft/drum (with deviations early on as different aeration strategies were tested).

Trial 2 consisted of corn stover alone or with a small amount of cracked corn grain addition. This trial was conducted over a period of six weeks at a moisture content of 50% (total weight basis) with four aeration rates. Aeration was done once each day.

Trial 3 consisted of the same stover and corn treatments but at 60% moisture and aeration at two rates. Air was supplied with higher frequency than in Trial 2, twice for the low rate and three times each day for the high rate. This was done in an attempt to increase the amount of time the compost was exposed to oxygenated air so that penetration to micropores would be enhanced.

Trial 4 was done on a larger scale primarily to address whether corn stover particle size reduction was necessary for composting. Finally the lessons learned from these trials were used to construct a larger pile equipped for heat recovery. Results are displayed in the accompanying tables. [Editor’s Note: For a copy of the tables, please contact NCR-SARE at 1-800-529-1342 or ncrsare@umn.edu.] Since successive trials are results driven from lessons learned from the previous ones, some results are discussed after each trial description. Most discussion is done in summary manner along with personal interpretation of the implications of the trials after all the results are disclosed.

RESULTS AND DISCUSSION
Trial 1
The aeration rate chosen for the first trial was arrived at as follows. If a 30-gallon compost drum containing 13.2 lb of shredded corn stover was expected to degrade at a rate of 25% over a six-week period, 3.3 pounds of oxygen or 215 cu ft of air would be needed over six weeks. The daily air volume requirement would be about 5 cubic feet. A small addition (2 pounds) of cracked corn was used on the assumption that the low bulk density of the stover with which it was mixed would make it less likely that oxygenation would limit the composting rate and the degradation rate of the grain would be easier to determine. Urea was added to evaluate the effect of lowering the C:N ratio. Although a nitrate form of nitrogen was preferred, urea was one of the most readily available sources of nitrogen in our area. The additional treatment of 10 pounds of cracked corn to the stover was done as an afterthought when the response to the 2-pound addition was so small.

After a few days, the strong ammonia odor emanating from the two treatments with urea addition provided assurance that there was ample urease activity for ammonification to occur (no attempt was made to quantify the ammonia loss through the breather vent, but it cleared my sinuses). Although unreplicated treatments cannot allow undeniable conclusions about responses to treatments, compost temperatures did not respond substantially to grain addition or urea addition. Temperature monitoring and aeration were discontinued after just 12 days because falling ambient temperatures made adequate temperature maintenance in the compost too difficult. The resumption of aeration after ambient air temperatures had warmed the following May stimulated composting to some extent especially where the higher grain quantity was added to stover.

The experience with the ammonia loss from the urea treatments reinforced the previously held assumption that lowering the C:N ratio for a material like corn stover, which has low degradability, with something like inorganic fertilizer nitrogen, which very quickly becomes available and is very easily lost in gaseous form, is not very straightforward. Ideally the availability of the nitrogen would coincide with its requirement by the organisms degrading the more recalcitrant material. Periodic small additions of fertilizer nitrogen might be matched better with the requirements if there was a good means to get it into the interior of the compost pile. The other point worth mentioning, although it was not addressed in this study, is that the widely suggested optimum C:N ratio of 25 to 30 is an average. Though it may be a good average for much of the biomass that is often composted, the range of optima might run from something close to 10-12 if sugar was composted to 200-400 for something like Western Red Cedar sawdust. Because of the relatively low degradability of corn stover, the high C:N ratio (not determined but assumed to be close to published values of about 50) may not benefit from N addition. The much higher degradability of corn grain might require that it be closer to the suggested optimum, which coincidentally it is (corn grain with 8% protein content should have a C:N ratio of about 30).

On June 1 the compost was removed from the drums. Both treatments with urea addition still smelled very strongly of ammonia, and the fresh appearance of the stover gave the impression that adding the urea was like adding a preservative. The treatment with 10 pounds of corn grain gave somewhat the same impression in appearance but smelled more like silage indicating inadequate oxygen supply, perhaps occurring after aeration was discontinued. Much of the grain seemed to have degraded. The stover alone and stover with 2 pounds of grain looked better decomposed than the others. The treatments without urea revealed that to a large degree a drying front had moved moisture from the interior of the compost and condensed it on the sides and top. Although an aeration rate of 5 cu ft/day was targeted it proved difficult to attain. So the aeration rate in this trial was probably much higher in an attempt to insure that degradability rather than oxygen was the rate-limiting factor during composting.

Trial 2
Since aeration was not very well controlled in the first trial and probably contributed to drying the compost more rapidly, stover alone and stover with 2 pounds of cracked corn was used in a second trial at four aeration rates. The mixtures were blended with water to 50% moisture content before placement in the drums. Initially, 19.5 lb (air dry) of ground stover was weighed into a tub for mixing. The moisture content was estimated to be about 13%, which meant 17 lb was dry matter and 2.5 lb was water. Another 14.5 lb of water was added and the stover mixed until the water was absorbed and blended in fairly uniformly. The first four drums consisted solely of this. Into the second four was added 2.35 lb of cracked grain (approximately 2 pounds dry matter) and an additional 1.65 lb of water before mixing.

During the first two weeks of the trial further attempts at continuous aeration resulted in failure to achieve consistency when the system was unattended. Since the pressure regulator did not maintain a constant pressure from one aeration to the next at such low pressures the aeration strategy was changed beginning on Day 15. Aeration during the remainder of the trial was done manually using a clock to determine air delivery and shutting off valves successively beginning with the lowest rate. Air delivery during each compressor cycle is known and doesn’t change with second stage pressure regulation. Each cycle of the compressor delivered 5.45 cubic feet of air and all eight drums were initially receiving air then each got 0.68 cubic feet/cycle. The time at the start of aeration and the cycle time was recorded so that the low rate valves were shut off after 3.3 cycles after all drums had received 2.25 cubic feet each. Then the remaining six drums were aerated for another 2.48 cycles receiving another 2.25 cu ft/drum and the second rate valves were turned off. The remaining four got 1.65 cycles worth of air for another 2.25 cu ft/drum before the third rate valves were shut off. Finally the remaining two drums got 0.82 cycles worth of air at which point aeration was complete. In this way the pressure could vary from one day to the next, but as long as the number of compressor cycles was used to meter airflow it did not matter. Since the compressor cycle time did not change as valves were closed during the initial aeration, increases in static pressure were not considered to be an issue. It was also assumed that since no significant air leaks were noticed in the air delivery manifold that there weren't any. The aeration rates beginning the third week of the trial were 2.25, 4.5, 6.75, and 9 cubic feet/drum per day.

Four times toward the end of the trial operator memory error caused excessive aeration. The temperature responses to these mistakes were instructive, revealing that aeration at the second rate was probably adequate for the stover alone but inadequate even at the highest rate where cracked corn had been added. After the temperature spikes following the excessive aeration the temperatures settled back into their pattern quite quickly when the normal aeration regime was resumed. (We often learn as much from our mistakes as from our intended purposes.) On Day 42 the prescribed quantity of air was added in split additions over multiple times during the day as indicated in the table. On Day 43 normal aeration was resumed, but temperatures were measured at multiple times during the day. On Day 44 air was supplied continuously. The data from Days 42 and 43 suggest that there was a temperature bump following aeration that subsided by the next day when aeration was normally done again. This temperature bump was missed when temperatures were measured only once each day just prior to aeration. It also showed that there was likely some benefit from adding air at greater frequency. Continuous aeration on Day 45 also gave further evidence of the advantage of continuous or more frequent aeration. The amount of air supplied beyond one pore volume per day does not seem so important as the amount of time that oxygenated air flows through macropores allowing time for diffusion into micropores.

The trial was terminated after 45 days. The drums were opened up for examination. The puzzling temperature response pattern in the stover with grain treatments became somewhat less puzzling as the compost was removed from the drums. The interior of the compost mass had become very dry, more so where there was grain added. Presumably this is because of the higher temperatures initially attained because of the grain. The top and bottom halves of the compost in each drum were removed and sampled separately to see whether there was a noticeable dry front due to aeration, which would leave the bottom drier. This did not seem to be happening to a noticeable extent. Samples from the interior of the compost mass and along the drum wall were also collected separately for moisture analysis (see Trial 2 moisture data in table).

The weight of compost remaining was determined by subtracting the empty drum weight from the weight with compost followed by a moisture content determination for calculation of the quantity of dry matter remaining. Compost from each drum was emptied into a tub for mixing and sampling. After blending the compost in the tub by hand, eight grab samples from each tub were combined and mixed in a small pail from which a 200-gram subsample was taken for dry matter determination. Samples were air dried in a makeshift solar dryer to remove most of the moisture. Several of these samples were then oven-dried at about 135 F for several hours. The moisture content values of all samples dried in the solar dryer were further adjusted based on these samples. (Dry matter determination is usually done based on higher temperature than this (175 F), but because the fire hazard in our kitchen oven was already high enough I settled for an approximation at a lower temperature.)

The wet compost along the drum wall and on top of the compost mass suggested that as the compost heated and evaporated water the steam condensed on the cooler surfaces along the drum wall and top surface. My interpretation is that where grain was in the mix this may have happened over a very short period early in the trial leaving the compost mass interior too dry to compost well resulting in lower than expected temperatures throughout the trial. This implies that keeping the compost at an optimum moisture content without turning may be very difficult. Maintaining the compost within a reasonable range might be challenging enough. Lacing PEX lines for heat recovery throughout the pile might help in this regard to slow the loss of moisture by providing cooler surfaces that condense steam in the interior of the pile where the moisture should ideally remain.

The Universal Heat Loss Equation (see footnote in Larger Trial data table) was used to estimate daily heat output and the fraction of heating value remaining after 45 days with an assumption of R-24 for the drum insulation value and an average temperature rise of about 4 degrees following aeration as indicated on the Day 43 temperature data. Tallying up the daily values and dividing by the total estimated BTU content (17.35 lb DM * 7000 BTUs/lb) gives a value of 91.4%, which is in the ballpark with the remaining dry matter remaining determined by weight difference. It was mentioned earlier that 25% mass reduction could reasonably be expected over a six-week period, but the reader is reminded that higher temperatures would be required than could be maintained in these screening trials using small drums.

Trial 3
Trial 3 treatments were identical to Trial 2 treatments with three exceptions: 1) The initial compost moisture content was 60% rather than 50% as in trial 2. 2) Two aeration rates, 5 and 10 cu ft/day, were done using split applications at multiple times per day. The low rates were split into two times about 8-10 hours apart. The high rate was split into three times about 8 hours apart. Each aeration consisted of filling the compressor to 125 pounds (2.67 cu ft tank) and letting it slowly bleed into the manifold. For two of the daily aerations all eight drums got air. For the last aeration of the day the valves to the drums getting the low rate were closed. When all eight drums were aerated they each got 2.5 cu ft. When only four were aerated they each got 5 cu ft. 3) Duplicates of each treatment received a water addition (4 ounces) once each week. This trial continued for about 30 days, into the middle of October when temperatures declined and made it difficult to maintain adequate composting temperatures in the drums. The response to water addition was noticeable only after the first addition one week into the trial. The higher moisture content at the start of the trial may have been adequate to keep the compost from drying out so much during the course of the trial.

The higher temperatures and the larger temperature differences from ambient air temperatures suggest more optimum moisture content and aeration regime than in the earlier trials. This may have been responsible for the more pronounced responses due to grain addition as well.

Trial 4
A trial using a larger pile size (about 500 pounds) was established for several reasons; 1) to compare shredded (in a tub grinder) and unshredded stover, 2) to see whether compost temperatures in the drums were so low mainly because of the small mass, 3) to see whether the same means of aeration through a 1/2" water line into the center of the pile was adequate in a larger pile (i.e. to determine ease of aeration).

The stover was spread on a clean, concrete cattle yard. Water was added through a hose that had been calibrated for flow rate. This was done with less accuracy than could be done in the drum trials because the stover weight was just estimated, and there was some water runoff. However, accuracy was considered to be sufficient for the purposes of the trial. The piles were wetted to about 50% moisture content, and placed in a bunker made from small square cornstalk bales. A single compressor hose was connected to a PVC pipe tee to evenly supply air to both piles simultaneously. The PVC pipe was placed so that the end was just off the bottom of the pile and at the center. The compressor was regulated to about a 7.5 minute cycle time. The air coming out of the supply line had a velocity of about 8 ft/sec at the center as determined using an air volume meter. Through a 1/2" tube this should deliver about 40 cu ft/hr to the two piles. When aeration was done for about 5 hours/day it was comparable to the rates used in the screening trials done in the 30-gallon drums.

The larger piles demonstrated a critical mass for heat generation and retention adequate to elevate temperatures to a desirable level (as high as 170 F) at least given the warm ambient temperatures during the trial. Although the shredded stover may have initially heated somewhat faster than the unshredded, shredding the stover does not seem necessary for it to compost effectively. Since shredding is quite energy intensive it doesn’t seem warranted. The trial was continued for less than three weeks, which seemed adequate since it was mainly to answer some preliminary questions prior to conducting a more carefully controlled compost pile of larger size.

Larger Pile with Heat Recovery
A larger pile consisting of an estimated 1500 to 1700 pounds of unshredded corn stover was located in a south facing lean off the side of a barn. The lean was protected from weather except on the south side. The compost was made from cornstalks alone. Two large round bales of stalks were rolled out and wetted with a garden hose for about 40 minutes. The flow was calibrated so that the stalks would be at least 40% moisture content after this amount of time. The late December day was unusually mild (40s) when most of the water addition was done, but by the time the pile was put together temperatures were in the single digits. Consequently there was insufficient time to reach the desired moisture content (50 to 60%). The compost pile moisture content was estimated to be 40% initially. More water was added to the pile at a later time during the trial, bringing the moisture content to 50% (see table).

A bunker was built out of small square cornstalk bales with a wall height of 4-5 ft and a floor of loose stover spread about 6” deep. This bunker was then lined with a plastic liner. For aeration a 3-4’ section of 4” corrugated plastic drainage tile was placed in the center on top of the shallow layer wetted stover. A 3/8” air hose was fed into the tile and routed into the barn to an air source. The air source was a CPAP (Constant Pressure Air Pump), which has an intended use by people with sleep apnea. Wetted stover was added layer by layer with coils of ½” PEX water line placed between each layer until about 220’ of PEX water line had been added. The PEX line was routed from the pile through two insulated PVC pipes (supply and return lines) to a small room in the barn, a distance of about 20’. The compost pile lay against the barn wall (insulated by cornstalk bales). The inlet and outlet lines were fitted with thermometer ports just on the inside of the wall from the pile so that they were within 2’ of the pile and insulated. The plastic liner was pulled over the top and tucked into the edges. Bats of fiberglass insulation were laid across the top of the pile. This insulation layer was then covered with a plastic tarp. An antifreeze solution was pumped through the pile in a closed loop system consisting of a pump, flow meter, expansion tank, thermometer ports, and a heater core fixed with a fan for heat extraction. Since heat recovery was measured by flow rate and temperature differential at the thermometer ports adjacent to the pile, for purposes of the experiment it did not matter whether the recovered heat was lost in the plumbing leading to the room being heated or in the room itself.

Because some of the stover had cooled to freezing or below at the time the pile was assembled a 1000 watt inline (radiator hose) engine heater was included in the loop long enough to provide heat to the pile until it was warm enough to generate its own heat. Although the RV antifreeze solution that was used indicated that it was effective to –50 F, it froze to slush even in the low teens. Since the month of January, 2009 never had temperatures above freezing in northern Iowa, it was never warm enough to thaw the slush from the system to make it possible to pump it until February.

The heater worked well so that within a couple days the pile was generating its own heat. The CPAP was then turned on to supply air continuously at an estimated rate of 1000 cu ft/day (estimated by the air volume meter at the end of about 30’ of hose). This air supply rate is similar on a mass basis to the higher rate used in the screening trials that were conducted in the drums. No attempt was made to extract heat until the compost temperature had risen sufficiently. After a couple days the temperature was 144 F. The data table indicates when the pump was turned on or off and the resulting pile and antifreeze temperatures. After the pump was turned on to extract heat, the temperature of the antifreeze solution fell quickly and remained quite low soon after starting the pump. After about 24 hours of pumping, the water temperature had dropped from around 60 F to the mid 30s with accompanying compost temperature in the 80s. The pump was shut off. After aerating for several days with no temperature response in the compost, warm water (115 to 120 F) was added to the compost to increase the moisture content and use the heat from the water to elevate the compost temperature. This elevated the temperature to 106 F several hours later, but by the next morning it was down to 90 F. Aeration for an additional 24 hours elevated the temperature to 128 F. A further 8 hours brought it up to 144 F later in the day. Subsequent heat extraction cycles were terminated before the compost temperatures had fallen to critically low levels, resulting in faster recovery times after heat extraction was stopped.

Considering the estimated heat capacity of the pile, the rapid drop in temperature when so little heat was extracted was puzzling. To see how fast the heat from the pile was lost to the surroundings after stopping aeration, the CPAP was shut off on Day 8 following a heat extraction cycle. By the next morning the pile temperature was at 145 F where it stayed for several days. It seemed, because of the very large temperature differential (100 to 150 F) over such a long time with no temperature drop in the pile, that aeration must have been occurring just by convection through holes and seams in the plastic liner without any assistance from the CPAP.

After composting for about 2 weeks the insulation and plastic was removed from the top of the pile. The surface 8-12” of compost was very wet. A moisture meter indicated dry conditions in the interior of the pile. A grab sample was taken from about 15” deep in the pile. It was very dry. The same moisture migration that was problematic in the screening trials using the 30-gallon drums was happening on a larger scale.

An estimate of the amount of heat that could reasonably be expected from the pile was made. The surface area of the pile (8’wide *10’long * 4’deep) was about 320 sq ft. The bulk density of the pile was about 5 lb/cu ft. The temperature difference between the compost and ambient air (delta T) after the pile was up to working temperature ranged from 80 F to 150 F during evaluations. The insulation value of the cornstalk bale bunker is hard to gauge but probably greater than R-20 based on published data for straw bales. Using the same assumptions as for the screening trials, about 25% of the 1500 lbs might reasonably be expected to degrade over a period of six weeks, about 9 lbs/day. Using an assumption that the energy content of the stalks is about 7000 BTUs/lb, the compost should generate 63,000 BTUs/day or 2600 BTUs/hr. Assuming the insulation value of the compost pile envelope averages R-20, the heat loss through the walls of the pile would be 1500 BTUs/hr, roughly 60% of the heat generated. This heat requirement to maintain the temperature of the compost mass under these extreme temperature differences seems to leave very little left for space heating in a pile this size. The pile would have to be much larger to get a reasonable amount of heat even for a small room.

SUMMARY DISCUSSION
Since I have already abandoned scientific rigor by not replicating the trials, I will follow a discussion based closely on the data with personal interpretation and speculation based on my observations, but hopefully based on the project results.

To summarize, there was no indication that the addition of nitrogen as urea was beneficial to the rate of composting corn stover and may have created an unhealthy environment for the organisms because of high ammonia levels. The addition of cracked corn resulted in greater heat output when moisture content and aeration rate were adequate. The stover composted comparably whether unshredded or shredded. This is important since shredding is an energy intensive and expensive operation. The stover proved easy to aerate because its low bulk density results in high porosity. Early intentions to optimize moisture and temperature by identifying particular levels were abandoned when it became apparent that simply maintaining certain ranges was challenging enough in static piles. The larger pile could maintain temperatures close to 150 F without heat extraction, but temperatures fell quickly with a modest heat removal rate. The main cause is probably low energy density, resulting from low bulk density and low degradability, and migration of moisture out of the interior of the pile to cooler exterior surfaces where it condensed and collected. Although the addition of cracked corn to the stover resulted in greater heat output in the screening trials, the rapid loss of moisture from the interior of the compost made determination of grain degradability difficult.

The consequence of low energy density (a consequence of low bulk density and low degradability) is that more volume and surface area is required for a given heat output rate, and more of the heat generated is lost because of the high surface area of the pile. Corn stover piled loosely as it was in the compost piles has a bulk density of about 5 lb/cu ft. Stover that is tightly compressed into a bale still has very low bulk density, about 8 lb/cu ft. The large volume of trapped air that gives dry corn stalks very good insulation value contributes to their low energy density. Even wet cornstalks laying on the soil surface result in much cooler soil beneath than where the soil surface is bare of residue.

Available energy density and porosity for aeration need to be balanced. An oxygen requirement that is too high for the aeration capacity will obviously result in anoxic conditions. That is one of the roles of bulking agents in composting. A dense, energy rich biomass like corn grain is more ideal in the sense of decreasing the surface to volume ratio but would require tumbling to provide adequate aeration if they comprised a large fraction of the compost. Continuous flow tumbling reactors that are commercially available for livestock mortality might be suitable for corn grain or DDGs with very little bulking agent addition. The most straightforward way to assess the heating potential of grain or DDGs in these types of reactors without reinventing the wheel would be to cooperate with an owner or vendor of one of these to modify it for heat recovery. Short of using a tumbling reactor, varying rates of inclusion of a more degradable constituent like grain with stover in static and aerated piles would identify the limits of aeration capacity to optimize the mixture.

Degradability is also important to match the composting rate to heating season requirements if a single pile is used. But it is also important in regards to the amount of heat available for space heating while leaving enough in the compost pile to compensate for heat lost to the surroundings in the process of staying hot enough to compost. If composting could be done inside the space to be heated, heat loss off the surface of the pile could be meeting space-heating needs rather than simply lost, as it would be if the pile were located outside. That is probably normally not possible for much the same reason that people don’t keep pet elephants in their house more often. There usually isn’t enough extra space, and it would smell bad. Although there have been instances where greenhouse operations not only have the space but can also make use of the carbon dioxide evolved from the compost to boost plant growth.

A more degradable material would provide a greater fraction of its heat for space heating but may not be sustained long enough into the winter to be tailored to heat requirements. The data from the drum trials show that controlling aeration rate could be a mechanism for controlling heat output. Including some grain is a means of increasing the energy density of the pile to generate more net heat for space heating. If that idea were taken far enough, silage harvesting the entire corn plant as is done for silage fed to cattle, might provide the needed quantity of grain along with sufficient bulking material from the stover. Silage is normally packed as a means of squeezing out the air to avoid spoilage (a.k.a. composting). If instead of packing an aeration mechanism was provided, the plant material would compost instead of (or to a greater degree than) ensile. The farm equipment for making silage already exists in abundance in most farming areas. Silage making would also preclude the need to add much water since it is harvested with ample moisture to be within the optimum range for composting. A serious caveat regarding the idea that corn silage could be composted successfully simply by aerating it is that for grain to compost well the kernel’s seed coat would probably need to be broken. Since this does not happen to any large extent during chopping, some means of scarifying the ears to damage the kernels might be necessary.

Another consequence of low bulk density is low thermal conductivity. These trials show how quickly the antifreeze temperatures fell in relation to pile temperatures when circulation through the compost pile began. This suggests that the low thermal conductivity within a compost pile may make providing heat at a comfortable temperature difficult even if the quantity of heat available is adequate. There may be good potential for a compost pile to supply heat that could be concentrated to usable levels by a geothermal heat pump without needing a large ground loop. However, if thermal conductivity alone were the main problem the pile temperature would fall only very slowly if at all when such a modest amount of heat was extracted. The fact that it fell so quickly suggests that either the energy density of corn stover alone is too low to be optimal, which is true, or that the interior of the pile was drying out just as it did in the drum trials. The samples pulled from the larger pile verified that this was also happening.

Moisture migration out of the center of the pile is one of the biggest challenges to overcome in a static compost pile. Although this would not be a problem in a turned pile the goal is to minimize the labor and energy inputs enough to make composting for heat recovery worth considering by having a means of heat removal embedded in a static compost pile. Using the PEX water loop to recover heat was initially considered to be a good means of heat recovery. In reality it (or something akin to it) may be essential to slow the flux of moisture out of the interior to the surface. A continuous flow of water through the lines that was varied in response to heat requirements and pile temperature would slow moisture loss from the interior. A means of recycling water back into the interior to replenish loss would eventually be necessary. Because heat extraction cooled the pile so quickly in this study, cooling through the PEX lines could not be done enough of the time to determine how well cooling kept the compost wetter around the water lines.

If not constrained by time and budget further work on this project would have continued its focus on the static pile, increasing the energy density of the composting material by inclusion of grain, distillers grain, food waste, or perhaps a more nitrogenous cellulosic biomass harvested at an earlier maturity to combine with corn stover to increase the energy density of the compost mixture. It should be kept in mind that the more energy rich constituents mentioned, with the exception of food waste, are more likely to have competitive markets as feed materials. This is another reason to continue to focus on corn stover or some other low cost residue as the base constituent for composting.

PROJECT IMPACTS
Crop residue in the Corn Belt may play a big role in the near future as the biofuel and chemical industry develops to make use of cellulosic biomass sources to meet the mandated renewable fuel goal. This future is frought with dangers and opportunities. The hazards are mainly to soil quality. Crop residue is so important for protecting soil from erosion and maintaining organic matter at levels conducive to high productivity. Some of the proposed conversion platforms would leave no organic residue for ultimate return to the soil. As producers consider the different markets for their crop residue market forces may encourage them to make unwise choices that result in near-term profits that externalize costs to future generations. Alternative markets and uses that will return nutrients and organic residue to fields that produce the biomass need to be emphasized and developed so that they are more attractive choices. This project fits into that theme since it would be an alternative use of a locally grown and abundant product that when used locally would permit easy recycle of both nutrients and organic matter back to the fields in a manner in keeping with sustainable and organic principles. At the same time, if successfully developed, it could reduce or preclude the need to purchase fossil fuel imports for heating fuel.

Although SARE projects like this would ideally result in turnkey ideas that could be readily employed on the farm, this is a project that hopefully contributes something to a concept that may ultimately be engaged on the farm but meanwhile requires more development. The ultimate goal that this project worked toward is to develop a composting system that is easy enough to use for space heating to be practical so that it is a realistic alternative to combustion of the biomass.

OUTREACH
The initial outreach effort was to make a poster presentation at the SARE Producer Conference at Oconomowoc, WI to elicit ideas to incorporate into the project. Although the project got exposure there, it proved to be a venue where people were trying to absorb as much as possible of the myriad projects represented. It was easy enough to get overwhelmed there without doing extra credit thinking. I got better feedback from occasionally bringing the project up amongst my peer producers that I know well and could engage periodically for their thoughts. No other means of outreach was planned until I could gauge the degree of success the project would have. My conclusion is that it was not ready to go to press yet.

PROGRAM EVALUATION
After two SARE producer grants I still feel very grateful for the opportunity to do some on-farm research that I would not otherwise be able to do. My only comment to the Administrative Council has to do with more and better interaction that some projects might benefit from at the time they are proposed. The personnel responsible for assessing the merit of the proposal and the capacity of the person (people) submitting the project idea could help strengthen the project by making suggestions for changes or additions, or connecting like-minded people together. It isn’t nearly as expensive or time consuming to conduct a project in your head using your imagination as it is to carry it out in real life. The more thought that goes into the project plan before it is launched the better the results will be.